How to Track Unique Objects Passing Through a Gate at High Speed?

[oliverrenner]

hello,

i need to uniqly identify up to 10 (or maybe more) different objects passing a 4x4m big gate with 200km/h speed.
right now we use an IR transmitter and reciever, but the problem is, one can hardly say for sure if the transmitter 100% passed though the gate, or just next to the receiver.
spontaneously i thought about a magnet and a conductor loop, but how can i differ between the magnets, since the weight should be the same of each.

does anybody have another smart idea ?

 

[anorlunda]

Can the objects carry an RFID chip? If so, then a variation of the EZ pass system used here to collect tolls from cars on the highway might work. In Florida, they also use cameras to recognize license plates at full highway speeds.

 

[oliverrenner]

yes they could. we also have power supply on our objects.
just rfid is quite expensive i think.

 

[anorlunda]

In Florida, they also use cameras to recognize license plates at full highway speeds.

 

[meBigGuy]

one can hardly say for sure if the transmitter 100% passed though the gate, or just next to the receiver.

I don't quite understand this. Is there an IR beam across the gate, and "things" pass through it breaking the beam? Or is there a transmitter on the object and it hits the receiver? How are you uniquely identifying the 10 objects now?

How constant it the speed? What range of speeds? Always 1 object at a time? Can you do something like use multiple magnets in unique physical patterns that let you identify the object when they pass a coil?

Maybe a drawing of the physical requirements/environment --- I'm pretty slow sometimes.

 

[Baluncore]

If you have two detectors then the sequence of detection will identify direction of travel.
Once detected by the first sensor, detection by the second will confirm passage through the gate.

 

[oliverrenner]

In Florida, they also use cameras to recognize license plates at full highway speeds.

yes that's an option, but i think its an quite expensive setup and difficult since the objects (to clarify we are talking about multirotor copters) move in more dimensions than a car on a road and video recognition tracking might get more difficult because of that.

I don't quite understand this. Is there an IR beam across the gate, and "things" pass through it breaking the beam? Or is there a transmitter on the object and it hits the receiver? How are you uniquely identifying the 10 objects now?

How constant it the speed? What range of speeds? Always 1 object at a time? Can you do something like use multiple magnets in unique physical patterns that let you identify the object when they pass a coil?

Maybe a drawing of the physical requirements/environment --- I'm pretty slow sometimes.

no, no, not slow. thank you for taking the time to think about my problem ! i am happy to provide more information!
- right now we have an IR transmitter on our object, which sends a unique id to an ir reciever inside the gate.
- speed is not constant at all, varys from 20 to 200km/h
- multiple objects passing at the same time is possible
- that's a very interesting idea with multiple magnets, i am just not sure about multiple objects at the same time...
but i know the topic is frustrating when one tries to find a perfect solution.

If you have two detectors then the sequence of detection will identify direction of travel.
Once detected by the first sensor, detection by the second will confirm passage through the gate.

i get the sequence of the detection leading to a direction of travel, but how can it confirm passing a gate ?

 

[Baluncore]

i get the sequence of the detection leading to a direction of travel, but how can it confirm passing a gate ?

The gate is between the two detectors.

 

[oliverrenner]

The gate is between the two detectors.

yes, but the ir is to inaccurate because of reflections.. the sensor would even trigger if a transponder passes half a meter behind him.

 

[Nidum]

Dumb transponders as used on shop doorway anti theft systems .

Multi light beams plus detectors on 2D grid as in early HP type touch screens .

Active transponders .

Acoustic detection by noise .

Acoustic detection by ultrasound .

Radar detection .

Use flat bed or side bed detection instead of gate - detect passing shadows .

Fit each item with a transponding GPS chip as used for following bird flight et al .

 

[Baluncore]

yes, but the ir is to inaccurate because of reflections..

The accuracy required is unspecified. Why are you using multipath ir ?
The only specifications we can guess at is that you are running a race with up to 10 vehicles, maybe motorcycles. Please fully specify the application. We should not have to guess.

 

[meBigGuy]

It really wastes a lot of people time when you don't accurately describe your requirements and environment.

so you are drone racing and want to score/time, or report drones passing through a gate. Multiple drones can pass at once.

I'm still not clear on the sentence "just next to the receiver.". You mean a drone is hovering at the receiver?

It seems you also have issues with drones blocking the IR reception of other drones.

 

[oliverrenner]

It really wastes a lot of people time when you don't accurately describe your requirements and environment.

so you are drone racing and want to score/time, or report drones passing through a gate. Multiple drones can pass at once.

I'm still not clear on the sentence "just next to the receiver.". You mean a drone is hovering at the receiver?

It seems you also have issues with drones blocking the IR reception of other drones.

sorry, its true i should have told it from the beginning.. i thought its not that important, my bad.
its about drone racing (multiple drones at once) and we want to track times (only counted if passing through the gate, premium version would be also recognizing passing NOT through the gate).
next to the receiver was ment that the drone might pass outside of the gate, but close enough that the ir signal is received from the receiver.
until now we had no problem suggesting that drones block other drones ir signal.

The accuracy required is unspecified. Why are you using multipath ir ?
The only specifications we can guess at is that you are running a race with up to 10 vehicles, maybe motorcycles. Please fully specify the application. We should not have to guess.

we did not build the current solution, it just works this way.
i hope i specified it with my response to meBigGuy.

 

[anorlunda]

Does it have to be automated? Couldn't you just take a video with a cell phone and have actual humans look at it to keep score?

By the way, I saw an article yesterday about how the NFL is tracking football players with inches of accuracy using RFID chips in their shoulder pads. You said that RFID is too expensive, but have you really investigated the costs?

If a cell phone video is too cheap, and RFID too expensive, what is your budget? You haven't said much about your actual requirements. We guess at ideas, you say no. It's time for you to write a real list of requirements including cost.

 

[oliverrenner]

Does it have to be automated? Couldn't you just take a video with a cell phone and have actual humans look at it to keep score?

By the way, I saw an article yesterday about how the NFL is tracking football players with inches of accuracy using RFID chips in their shoulder pads. You said that RFID is too expensive, but have you really investigated the costs?

If a cell phone video is too cheap, and RFID too expensive, what is your budget? You haven't said much about your actual requirements. We guess at ideas, you say no. It's time for you to write a real list of requirements including cost.

yes automated is a must... otherwise we can go back to just using colored copters and people with watches at the gates. we need real-time data.
i heared about the NFL football player tracking.. i will investigate how they do it.. this sounds quite promising that's true.
costs is relative.. rfid systems i know (which might not represent the whole market) cost 3000-5000 USD, where the IR system we use now costs 300USD.
it would be nice to provide a solution below 1000USD, but of course if there is non, the theoretical budget will be increased until there is.
all other requirements are stated in post #13.. its about multiple copters racing in rounds through gates, and we want to track unique copters passing through a gate in real time.